Pressurized Oxygen Blown Entrained-Flow Gasification of Wood Powder

Article pubs.acs.org/EF

Pressurized Oxygen Blown Entrained-Flow Gasification of Wood Powder Fredrik Weiland,*,†,‡,§ Henry Hedman,† Magnus Marklund,† Henrik Wiinikka,†,‡ Olov Ö hrman,† and Rikard Gebart‡ †

Energy Technology Centre in Piteå, Sweden Division of Energy Science, Luleå University of Technology, Sweden

‡

ABSTRACT: In the present study, an oxygen blown pilot scale pressurized entrained-flow biomass gasification plant (PEBG, 1 MWth) was designed, constructed, and operated. This Article provides a detailed description of the pilot plant and results from gasification experiments with stem wood biomass made from pine and spruce. The focus was to evaluate the performance of the gasifier with respect to syngas quality and mass and energy balance. The gasifier was operated at an elevated pressure of 2 bar(a) and at an oxygen equivalence ratio (λ) between 0.43 and 0.50. The resulting process temperatures in the hot part of the gasifier were in the range of 1100−1300 °C during the experiments. As expected, a higher λ results in a higher process temperature. The syngas concentrations (dry and N2 free) during the experiments were 25−28 mol % for H2, 47−49 mol % for CO, 20−24 mol % for CO2, and 1−2 mol % for CH4. The dry syngas N2 content was varied between 18 and 25 mol % depending on the operating conditions of the gasifier. The syngas H2/CO ratio was 0.54−0.57. The gasifier cold gas efficiency (CGE) was approximately 70% for the experimental campaigns performed in this study. The synthesis gas produced by the PEBG has potential for further upgrading to renewable products, for example, chemicals or biofuels, because the performance of the gasifier is close to that of other relevant gasifiers.

■

INTRODUCTION It is estimated that the total world energy demand will increase by 40% over the next 20 years, and one of the fastest growing sectors is the transportation sector.1 Second generation biofuels based on nonfood biomass is a priority in Europe due to high level political decisions on the European level2 aimed at reducing the emission of greenhouse gases and increased security of supply without negatively impacting the food supply. One of the goals in the national energy strategy of Sweden is to make the vehicle fleet independent of fossil fuels by 2030.3 Nonfood biomass can come from either agricultural residues or forest biomass. The main natural resource available for this in Sweden is forest residues that cannot easily be used by the forest products industry. In a comparison of potential processes for conversion of biomass into biofuels, it was found that thermal gasification of wood and black liquor followed by conversion of the resulting syngas into dimethylether (DME) had a competitive well-to-wheel efficiency.4 Hence, this has become one of the main routes being considered in Sweden and demonstrated in a European project where BioDME (DME from biomass) was produced, distributed, and tested in a vehicle field test.5 Techniques for syngas generation from biomass can crudely be divided in low-temperature (fluidized bed and fixed bed) and high-temperature (oxygen blown entrained flow) processes. Oxygen blown high-temperature entrained-flow gasifiers are designed to work in the slagging mode, which means that, as long as the gasifier temperature exceeds the ash slag fluid point, variations in the ash melting point are less problematic than in fluidized bed gasifiers.6 Entrained-flow gasifiers can potentially be used for biomass gasification with a feed of small particles or © 2013 American Chemical Society

bio-oil at large capacity, at high pressures, high temperatures, and with a short residence time resulting in a clean syngas with a very low tar content. The drawback with operating in the slagging mode is that the temperature is so high that the durability of the containment materials becomes a concern and that the recovery of heat in the hot syngas becomes more important for the overall efficiency of the process than with low-temperature gasifiers. The fluidized bed gasification technology is limited by the fact that the bed temperature must be safely below the ash melting point. This results in a tarrich syngas that must be further upgraded and cleaned6 before conversion to chemicals or biofuels. This adds complexity as compared to the entrained-flow process. Swanson et al.7 assessed biofuel production based on gasification with a fluidized bed and an entrained-flow concept. They concluded that the entrained-flow concept, due to its higher carbon conversion, resulted in lower fuel production cost, although the capital cost for the entrained-flow technology was about 20% higher than the fluidized-bed scenario. Further research and development of high-temperature entrained-flow biomass gasification, especially large-scale applications, is therefore of interest.7,8 Several gasification related parameters have to be investigated before conversion of biomass into biofuels can be commercialized. The syngas quality is of great importance for the design of downstream syngas cleaning and conditioning equipment. Quench water or condensate contaminants must be investigated to address proper quench water handling to avoid Received: November 5, 2012 Revised: December 23, 2012 Published: January 24, 2013 932

dx.doi.org/10.1021/ef301803s | Energy Fuels 2013, 27, 932−941

Energy & Fuels

Article

system was used, and relevant process values were stored in a database for post processing of the results from the experiments. The PEBG gasifier consisted of a ceramic lined reactor (0.52 m in inner diameter and 1.67 m in vertical reactor wall length) with a conical shaped outlet followed by a bubbling two-level water sprayed quench for syngas cooling and smelt/particle separation. The hot face ceramic material in the reactor consisted mainly of Al2O3 (63 wt %) and SiO2 (31 wt %). In total, five thermocouples (Type S, shielded with protective ceramic encapsulation) measured the process temperatures at different locations inside the reactor, placed with the tip coincident with the inner reactor wall. One thermocouple was mounted in the upper part of the reactor, three thermocouples were mounted at the reactor middle circumference (equally separated by 120°), and one thermocouple was mounted at the lower part of the reactor (just above the conical section). Unless specified otherwise, when referring to the process temperature in the text, it refers to the average temperature obtained from the three thermocouples mounted in the middle part of the reactor. To visualize the flame and the interior of the reactor, a water cooled N2 purged camera probe (see Figure 3) was installed in the top end of the reactor, slightly tilted toward the center axis of the reactor. The N2 purging flow through the camera was restricted by an orifice plate at the back end of the probe to determine the mass flow of N2 from the pressure difference over the orifice. The fuel powder was prepared by milling the considered fuel material in a hammer mill (MAFA EU-4B, not presented in Figure 2) where the sieve size in the mill was used to affect the resulting particle size distribution of the fuel powder. After milling, the fuel powder was pneumatically transported to two hopper storage tanks in the feeding system located above the reactor in the PEBG plant. The two hopper tanks (approximately 1 m3 each) were sequentially filled and alternately operated to keep the pressurized process in continuous operation. The accuracy of the fuel feeding rate in the mechanically based feeding system was found to be within ±3% (95% confidence interval). The accuracy was determined by repeatedly logging (at 1 Hz) the discharged weight of fuel from the fuel feeding system. This calibration check was performed for every new fuel type or fuel size distribution prior to the gasification experiment. For safety reasons, a (small) mass flow controlled (MFC) stream of N2 (via a Bronkhorst F203AI) was also transported with the fuel powder to keep an inert atmosphere within the fuel feeding system down to the burner. This flow mix of fuel powder and transport N2 was then centrally introduced in the top of the reactor. N2 could also be added as a distributed series of “shock puffs” into the lock hoppers to break potential powder bridges that could cause interruptions in the fuel feeding. The supplied amount of O2 used in the gasification process was controlled by a MFC (Bronkhorst F-203AI) and introduced in the top of the reactor via an O2 register located concentrically outside the fuel entrance of the burner. For process safety reasons, the flow rate of O2 was also measured with a secondary mass flow meter (Yokogawa Coriolis Rotamass RCCS31). Furthermore, it was also possible to introduce and mix N2 (via another N2 MFC, Bronkhorst F-203AI) with the O2 stream to simulate different O2 concentrations (from 21 to 100 mol %) in the O2 register. Prior to process start-up, the reactor ceramics was slowly heated (20%. λ=

ṁ actual ṁ stoichiometric

(1)

where ṁ actual and ṁ stoichiometric are the actually supplied O2 mass flow (kg/s) and the stoichiometrically required (for complete combustion) O2 mass flow (kg/s), respectively. This Article presents results from two different process conditions where the parameters oxygen equivalence ratio (λ) and mass median fuel particle size (d50) were varied. The considered process conditions were (a) λ ≈ 0.44 with d50 ≈ 100 μm and (b) λ ≈ 0.50 with d50 ≈ 200 μm. One replicate experiment at each process condition resulted in four experimental campaigns performed in the beginning of 2012 (named as January 12, February 14, February 16, and February 17 in Table 2). Gas and Particle Sampling. An overview of the syngas sampling system is shown in Figure 5. Syngas sampling was performed by letting a small slip stream of the produced syngas flow from the outlet syngas pipe (after the pressure relief valve) through a particulate removal equipment and a water cooled condenser. The particulate removal equipment consisted of a fiberglass filled trap followed by a glass microfiber disc filter. In the water cooled condenser, the cooling water inlet temperature was 10 ± 2 °C during the experiments. The dried syngas was then analyzed consecutively by a FTIR instrument (MKS

kg/h in all of the experiments, corresponding to approximately 210 kWth. The oxygen equivalence ratio (λ) was defined as the ratio between the supplied O2 and the stoichiometric O2 demand for complete combustion:

Figure 4. Particle size distributions for the considered fuel milled with 0.75 and 1.5 mm hammer mill sieve size, respectively. The corresponding mass median particle diameters, d50, were approximately 100 and 200 μm for hammer mill sieve size of 0.75 and 1.5 mm, respectively. 935

dx.doi.org/10.1021/ef301803s | Energy Fuels 2013, 27, 932−941

Energy & Fuels

Article

Figure 5. Schematic overview of the syngas sampling system. Multigas 2030 HS) and a micro-GC (Micro Gas Chromatograph Varian 490 GC with Molecular sieve 5A and PoraPlot U columns). The FTIR continuously detected and logged CO, H2O, CO2, and CH4 concentrations at 1 Hz, whereas the micro-GC detected and stored H2, N2, O2, CO, CO2, CH4, C2H4, and C2H2 concentrations every 4 min. An additional gas stream was also sampled from the syngas outlet pipe to determine the concentration and size distributions of particulate matter in the syngas by a Dekati low pressure impactor (DLPI). The detailed results and implications of the particle measurements are, however, considered to be out of the scope for this work and will be presented elsewhere. However, the obtained total concentration of particles was used for the mass and energy balances for one of the presented experimental campaigns (January 12). Quench Water Sampling. Samples were consecutively taken from the quench water outlet after the sedimentation tank for one (the January 12 campaign) of the four experimental campaigns. To determine the quench water solid content, the collected samples were vacuum filtered through a membrane filter (pore size 0.45 μm). One representative sample was sent for analysis of total organic carbon (TOC) content to estimate the carbon losses to the quench water. The amount of accumulated soot and char particles within the system, that is, in the sedimentation tank and quench water pool, was not determined. Mass and Energy Balances − Definitions. The total syngas flow was not measured directly. Instead, N2 was used as a tracer, making it possible to calculate the total mass flow of dry syngas from the N2 concentration in the syngas according to eq 2. This is possible under the assumption that all of the inlet N2 ends up in the syngas, without transforming into other compounds. Note that the fuel-N was neglected because of the low N-content in the fuel (0.10 wt % ds). The molar flow rate of syngas was calculated according to nsyngas = ̇

n ̇N2 c N2

CGE =

ṁ in =

(4)

ṁ fuel · cC 100

(5)

where ṁ fuel is the mass flow of dry fuel (kg/s), and cc is the dry fuel carbon content in wt %.

ṁ out = nsyngas · ̇

(cCO + cCH4 + cCO2 + 2·cC2H4 + 2·cC2H2) 100

· MC (6)

where ṅsyngas is the molar flow (kmol/s) of syngas, c is the concentration of each species in the syngas (mol %), and MC is the molar mass of carbon (i.e., 12.01 kg/kmol). The carbon conversion is defined as the ratio between the determined output and the input carbon flows from above. Energy losses derived from soot in the syngas and in the quench water outlet were estimated by assuming that the soot was pure carbon with a lower heating value (LHV) of 34 MJ/ kg. Note the oxygen and hydrogen balances could not be closed because the water yields were not accurately determined; therefore, only the carbon balance is presented in this work. Propagated measurement errors were estimated for some of the key figures related to gasification performance (e.g., CGE, mass and energy balances, etc.). The error estimations were calculated on the basis of a general formula for error propagation.12 The experimental error was estimated only for one of the experimental campaigns (January 12), and it was assumed that the experimental errors were similar for all four campaign dates described in this Article.

·100 (2)

1 ·∑ ci·LHVi 100 i

ṁ fuel · LHVfuel

where the subscripted ṁ represents the mass flow (kg/s) of fuel and syngas, respectively. LHV represents the lower heating value (MJ/kg) for fuel and syngas (subscripted), respectively. With respect to the carbon mass balance, the input and output carbon mass flows (kg/s) were calculated according to eqs 5 and 6, respectively.

where ṅ is the molar flow rate in kmol/s, and c is the concentration (mol %) of N2 in the syngas. The lower heating value (LHV) of the syngas was calculated as the sum of each components contribution to the heating value according to

LHVsyngas =

ṁ syngas · LHVsyngas

■

RESULTS AND DISCUSSION This study originates from the initial 100 h of operation. In the unlikely event of a syngas explosion, the resulting pressure would be sustained within the process equipment if the initial pressure does not exceed 2 bar(a). Therefore, the gasifier was operated at pressures ≤2 bar(a) in this work for process safety reasons. The effect of increased system pressure will be investigated later and presented elsewhere. The reactor camera enabled visual access to the reactor interior and contributed considerably to increased process safety; see examples in Figure 6A−D. It was a qualitatively

(3)

where the components i are H2, CO, CH4, C2H4, and C2H2. Lower heating values for each component were found in existing literature.10,11 The cold gas efficiency (CGE) for the process was defined as the ratio between the energy in the produced cooled syngas and the energy input from the corresponding fuel according to 936

dx.doi.org/10.1021/ef301803s | Energy Fuels 2013, 27, 932−941

Energy & Fuels

Article

presented in Figures 7 and 8, respectively. The graphs represent a typical experimental campaign in the PEBG pilot plant. The

Figure 8. Gas composition profiles for the experimental campaign January 12.

presented temperatures in this work were the measured temperatures of the thermocouples and not the real gas temperatures. The main error in the temperature measurements is mainly due to radiative heat transfer between thermocouple, the flame, and the surrounding walls. Generally, the top and middle temperatures along the reactor axis were similar, whereas the bottom reactor temperature always was approximately 60 °C lower. In the typical experiment shown in Figures 7 and 8, the reactor was electrically heated to approximately 1020 °C, after which the heater was removed and the reactor N2 purged to evacuate all possible O2 in the gasifier prior to start-up. After a first process start-up and approximately 30 min of operation, this particular experimental campaign (January 12) had to be temporarily interrupted and the reactor purged with N2. After the second start-up (time = 220 min in Figures 7 and 8), the plant was continuously operated for 6 h 50 min. Between time 220−305 min, the gasifier was operated with a λ at 0.50 to quickly reach the desired process temperature of ∼1200 °C

Figure 6. Photos (A−D) taken by the reactor camera from inside the reactor. The photos were taken at different times during the process. Photo (A) shows the electrical heating of the reactor; photo (B) shows the fuel feeding start-up; photo (C) shows process ignition; and photo (D) shows the gasification flame. The reactor outlet is visible as a black hole at the bottom of pictures (A) and (B).

valuable tool for operation, especially at process start-up, when the camera made it possible for the operators to actually see when wood powder was introduced to the reactor; see Figure 6B. It also gave a very fast response on the ignition phase as shown in Figure 6C, when O2 was supplied and ignited the pyrolysis gases. Furthermore, the camera made it possible to monitor the burner flame qualitatively; see example in Figure 6D. Process Characterization. Four experimental campaigns are described in this work. The process temperature profile (middle reactor temperature) and dry product gas profile as a function of time for one of the campaign dates (January 12) are

Figure 7. Process temperature profile (middle reactor temperature) for the experimental campaign January 12. 937

dx.doi.org/10.1021/ef301803s | Energy Fuels 2013, 27, 932−941

Energy & Fuels

Article

coherently equal gas data from the micro-GC. This, for example, occurs after approximately process time 400 min in the experimental campaign January 12; see Figure 8. Carbon Mass and Energy Balances. As described in the Method, there were four different N2 inlet flows, and it was found that the error in N2 mass flow had the individual largest contribution to the experimental errors of both the carbon mass and the energy balances. Hence, improving the accuracy of the trace gas mass flow would strongly increase the accuracy of mass and energy balances. The amount of particulates in the syngas was estimated to approximately 200 mg/Nm3, where the majority was assumed to be soot because of the black color of the deposits on the impactor plates. Analysis performed on the impactor plates with SEM/EDS (scanning electron microscopy/energy dispersive spectroscopy) showed that the deposits consisted mainly of carbon. Particulate matter in the syngas will be further investigated in future studies. Quench water sampling after the sedimentation tank (January 12) resulted in 65 mg TOC/dm3 quench water. The level of soot in the syngas and quench water had only a minor effect on the mass and energy balances and was therefore neglected for the experimental runs after January 12. Particulate matter also accumulated in the quench pool and in the downstream quench water sedimentation tank during all campaigns. However, the amount and composition of the accumulated matter were not determined in this work, but need to be studied in future experiments. The carbon conversions (Cout/Cin) for the four tests are presented in Table 4. The carbon conversions were

(middle reactor temperatures). A minor process temperature dip in an otherwise continuous temperature increase occurred when the operated λ was reduced to 0.44. Approximately 30 min after the change in λ, the temperature was again continuously increasing with approximately 20 °C/h to the end of the gasification test. This indicates that the system did not reach thermal equilibrium even after nearly 7 h of operation. The measured process temperatures should be considered as the ceramic wall temperature. The flame and the hot gases in the reactor are slowly heating the reactor wall after gasification start-up. The heating of the reactor wall is slow due to the inertia of the ceramic mass. Even though the process temperatures were continuously increasing in the described experiment (Figure 7), the concentration of the major syngas components (H2, CO, and CO2) seemed to stabilize approximately 30 min after the λ reduction; see Figure 8. This might indicate that the reactions in the gas phase are mostly dependent on the stoichiometry (λ) and the real gas temperature in the reactor. However, the CH4 content in the syngas showed a slight decrease, from approximately 2.0 to 1.8 mol %, as the process temperature increased. Unfortunately, the CH4 concentration at thermal equilibrium could not be determined in this work; neither was the content of higher hydrocarbons or tars determined in this work. There could potentially be some content of higher hydrocarbons and tars in the syngas, and this will be further investigated in future experiments. The ragged part of the curves in Figures 7 and 8, that is, from process time 470 min and forward, arose from irregular N2 addition to the fuel feeding system. Note that this is not a typical operating procedure for the PEBG plant, but was tested for this specific experimental campaign. Additional “shock puffs” were introduced to the second fuel hopper for improvement of the fuel feeding. Therefore, the gasifier temperature dropped every time additional N2 was added to the system. Moreover, the syngas was irregularly diluted with N2, which can be visualized by the irregular main gas compositions after process time 470 min in Figure 8. Table 3 presents the syngas compositions for all four experimental campaigns at stable condition. Stable gas compositions were determined when there were longer periods (>1 h) with

Table 4. Calculated Carbon Conversions for the Considered Experimental Campaigns

12 Janα

unit

7.7 ± 0.3 syngas LHV MJ/kg syngas mass flow kgb/h 69.9 ± 10.6 N2 content mol % 18.1 H2/CO 0.57 ± 0.04 N2 Free Dry Gas Composition H2 mol % 27.8 CO mol % 48.7 CO2 mol % 20.4 CH4 mol % 2.3 C2H4 mol % 0.1 C2H2 mol % 0.3 b

14 Feb

16 Feb

17 Feb

7.3 74.6 21.1 0.57

7.0 75.2 18.0 0.54

6.2 79.5 25.2 0.54

27.8 48.5 21.1 2.3 0.1 0.3

25.8 48.0 23.1 1.4 0.1 0.2

25.4 47.3 23.9 1.4 0.1 0.2

unit

12 Janα

14 Feb

16 Feb

17 Feb

Cout/Cin

%

103 ± 17

106

110

105

α

The experimental error was estimated only for one of the experimental campaigns (January 12), and it was assumed that the experimental errors were similar for all four campaign dates described in this Article.

approximately 100%, or actually slightly above (see Table 4). It is however impossible to have carbon conversions above 100%, and hence the conversion figures indicate that either the output carbon flow was overpredicted or the input carbon flow was underestimated. As mentioned earlier, the uncertainty in the syngas mass flow was large, which also was reflected in the estimated error of the carbon conversion (∼17%, see Table 4). Table 5 shows the energy balance for the four different campaigns. The thermal power from the fuel was approximately 210 kW for all cases. Of this, approximately 140−150 kW ended up as chemical energy in the cold syngas depending on the operating conditions. Thus, the cold gas efficiency (CGE see Table 5) was found to be approximately 70 ± 12%. It is worth mentioning that the process was not optimized in any sense, because process optimization was beyond the scope of the present study. The estimated error of the CGE includes the potentially overpredicted syngas mass flow. Furthermore, part of the energy generated in the gasification process was transferred to the quench water during cooling of the hot gas from the reactor. The quench water temperature was increased from approximately 8 °C (quench water inlet) to between 45 and 60 °C (quench water outlet). The reference

Table 3. Syngas Characteristics for the Four Experimental Campaigns dry syngas

carbon conversion

α

The experimental error was estimated only for one of the experimental campaigns (January 12), and it was assumed that the experimental errors were similar for all four campaign dates described in this Article. bKilograms of dry gas. 938

dx.doi.org/10.1021/ef301803s | Energy Fuels 2013, 27, 932−941

Energy & Fuels

Article

that soot and unconverted char accumulated in the quench where its energy content was not taken into consideration in this study. However, similar to the carbon mass balance, the Eout/Ein had a relatively large estimated error (∼12%, see Table 5), which also takes the potentially overpredicted syngas mass flow into account. Syngas Characterization and Utilization. Table 3 show the N2 free dry gas composition together with some dry syngas key figures. There were only minor variations in the dry and N2 free product gas composition among the campaigns with similar process conditions; see Tables 2 and 3. The major syngas components for process condition (a) were approximately (mol % dry and N2 free): H2 ∼28%, CO ∼49%, CO2 ∼21%, and CH4 ∼2.3%. Process condition (b) resulted in syngas composition of approximately (mol % dry and N2 free): H2 ∼26%, CO ∼48%, CO2 ∼24%, and CH4 ∼1.4%. The syngas heating value (LHV) was reduced for condition (b) as compared to (a), as expected, when the process was operated at a higher λ. An increased O2 addition to the gasifier, that is, higher λ, promotes the combustion of the H2, CO, and CH4 components and therefore consumes part of the chemical energy otherwise stored in the syngas. The LHV was also dependent on the amount of N2 in the syngas. The process benefits from minimizing the amount of inert gas (N2), not only because of the reduced dilution effect on the LHV, but also because of the decreased amount of energy that is required to heat the N2 to the process temperature inside the gasifier. In this work, the N2 addition to the O2 burner register had the purpose of a moderator to keep the burner temperature below an acceptable limit. However, alternative cooling of the burner will be considered in future work to decrease the N2 addition to the gasification process. Moreover, Table 3 shows that the H2/CO ratio was 0.57 and 0.54 for the process conditions (a) and (b), respectively. This was, however, only a minor change, which makes it hard to draw any direct conclusions, but the difference may still indicate that the combustion reactions of H2 and CO were promoted differently when more O2 was added. Moreover, the fuel particle size was smaller for process condition (a) as compared to process condition (b) according to Table 2. Therefore, it is

Table 5. Total Energy Balance for the Considered Experimental Campaigns In LHV fuel fuel sensible quench water inlet sensibleb total in Out LHV syngas syngas sensible soot in syngas TOC in quench water quench water outlet sensible heat loss total out Eout/Einc cold gas efficiency (CGE)

unit

12 Jana

kW kW kW

14 Feb

16 Feb

17 Feb

210 ± 8 0.1 −8 ± 3

210 0.1 −8

210 0.1 −8

210 0.1 −8

kW

202 ± 8

202

202

202

kW kW kW kW kW

150 ± 24 0.6 0.3 0.4 35 ± 3

151 0.3 n.a. n.a. 23

147 0.5 n.a. n.a. 31

137 0.5 n.a. n.a. 32

kW kW % %

8 194 96 71

8 182 90 72

8 186 92 70

8 178 88 66

± ± ± ±

1.6 24 12 12

a

The experimental error was estimated only for one of the experimental campaigns (January 12), and it was assumed that the experimental errors were similar for all four campaign dates described in this Article. bThe reference temperature was 18 °C. The quench water inlet temperature was approximately 8 °C, thereby the negative value of the quench water inlet sensible heat. cEnergy ratio: Eout/Ein.

state used in the calculations was 18 °C, and therefore the energy loss to the quench water was divided into two parts in Table 5: one (negative) inlet sensible heat and one outlet sensible heat. Convective and radiative heat losses from the reactor to the surrounding atmosphere (∼8 kW, Table 5) were estimated from the measured reactor shell temperatures and a heat balance calculation taking into account radiation and natural convection to the surroundings. The energy balance ratio Eout/Ein was close to ∼100% (Table 5). Actually, the Eout/Ein figures for all four campaigns were slightly lower than 100%. This might be explained by the fact

Table 6. Typical Gas Compositions (mol %, Dry Basis) for Different Gasification Techniquesa gasification concept thermal power input (MW) feedstock gasification agent gasification temp (°C) gasification pressure (bar(a)) Typical Gas Compositions (mol H2 N2 CO CO2 CH4 other H2/CO (H2 − CO2) (CO + CO2)

LHV (MJ/kgdry) CGE (%) a

PEBG pilot

cyclone

BLG

FB Värnamo

FICFB Vienna

EF coal

entrained flow